Real-time implementation of HMM-based chord estimation in musical audio

Taemin Cho, Juan P. Bello

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we implement a real-time chord estimation system based on the hidden Markov model (HMM) using chroma features. HMMs, with the Viterbi decoding algorithm, have proven a powerful tool for chord recognition in polyphonic music signals. However, the direct application of the traditional, non-causal, decoding approach is not the optimal choice for real-time processing, with limited memory capabilities and no access to future observations. We propose a system of buffers and a modified decoding process that approximates offline results while minimizing the system's latency.

Original languageEnglish (US)
Title of host publicationProceedings of the 2009 International Computer Music Conference, ICMC 2009
PublisherInternational Computer Music Association
Pages117-120
Number of pages4
ISBN (Print)9780971319271
StatePublished - 2009
Event2009 International Computer Music Conference, ICMC 2009 - Montreal, QC, Canada
Duration: Aug 16 2009Aug 21 2009

Publication series

NameProceedings of the 2009 International Computer Music Conference, ICMC 2009

Other

Other2009 International Computer Music Conference, ICMC 2009
Country/TerritoryCanada
CityMontreal, QC
Period8/16/098/21/09

ASJC Scopus subject areas

  • Music
  • Computer Science Applications
  • Media Technology

Fingerprint

Dive into the research topics of 'Real-time implementation of HMM-based chord estimation in musical audio'. Together they form a unique fingerprint.

Cite this